If you're referring to the well-established result that the ear perceives small level changes as changes in sound quality, then yes, that's absolutely true- as long as the levels are controlled and the switching is rapid. Otherwise, I'm unfamiliar with any published data indicating that 0.12dB (a curiously precise number gotten from where?) is perceptible when there is a gap between presentations (e.g., the typical audiophile cable swap). Nor do I see any connection to living-room music listening, where levels are anything but controlled.
Daniel,
The Fletcher-Munson equal loudness curves have been known to most of us for many decades!
They are the result od experiments with subjects listening to SOUND. ie they apply equally to live and recorded sound.
The variations with level are quite significant w.r.t small changes in FR due to cables - if such exist.
There are some very learned and knowledgeable people on here. Quoting of numbers to support ideas need to be well thought through if said ideas are not to be challenged!
The Fletcher-Munson equal loudness curves have been known to most of us for many decades!
They are the result od experiments with subjects listening to SOUND. ie they apply equally to live and recorded sound.
The variations with level are quite significant w.r.t small changes in FR due to cables - if such exist.
There are some very learned and knowledgeable people on here. Quoting of numbers to support ideas need to be well thought through if said ideas are not to be challenged!
Let's say you have a DUT that has 0.1dB more gain (or less loss) than another. It will be perceived by a sharp ear as sounding subtly clearer, not louder. If the volume control (or an attenuator) is not adjusted to keep levels the same, you have a confounding variable. This would be true if the DUTs are two wires with a small difference in DCR. The DCR forms a voltage divider with the load.
A 0.1dB overall level change is barely perceptible to a true golden ear, assuming a rapid A/B comparison. EQ is trickier to judge; a 0.1dB EQ change at 1kHz might be barely perceptible to the sharpest ears, 0.1dB at 20kHz, much less likely.
And I should add that a 0.1dB level or EQ difference between two channels may also be perceptible, possibly even moreso. That's one reason that it's so important to have RIAA stages have very tight matching channel-to-channel.
More.
It is both mechanical and electronic modifications to a highly sensitive speaker driver, and it is very much application specific on equipment requirements.
Normally, in such cases, the equipment is all bolted together and named "active" in order to avoid argument about optional components.
I think that the total system symmetry approach is Much better than using an equalizer to put gain on error in the hopes that two wrongs might make a right.
When people say they hear differences-between cables-surely they don't mean they hear 0.1 differences at 20Khz.
One point recordings are quite rare these days, but if you know where to look, you still will find some.
As a curiuosity Stephen Epstein, a long time producer of classical music at Sony Music, sometimes chooses to record with just one pair if possible.
As examples the whole Sony series of Sibelius orchestral works are one point recordings featuring B&K omni mikes.
Overly capacitive speakerwires are as you mentioned yourself, pretty heavy on sonic impact, try to read these words from MIT cables own homepage:
"When you connect an MIT cable up to your system, you will enjoy recorded music in an entirely new way. The background is black, increasing the contrast between the instruments and voices. A lifelike soundstage opens up with pinpoint imaging at any volume level. "
The one with the black background has always amused me a lot, as I´ve attended many many concerts, but nowhere was there any black background what so ever. Especially when listening to one point recordings with omni mikes, there should never be completely quiet, as people make noise, even if they try not to, and a whole orchestra makes even more. This is to be easily heard in a good system, but connecting the MIT´s it actually disappears, and blacknes appears instead.
Overly inductice cables as well as very stiff and mechanical resonant cables, I also think should be avoided.
Where did you get this number?
My own experience tells me that with complex signals, eg music, that quite large gain differences aren't usually heard as such. My understanding too is that the research into thresholds was done with single tones, which would make it easier to detect small changes.
Practically everyone who purchased a Creative Labs X-Fi Xtreme Music and then also went out and purchased an M-Audio Audiophile 192 (or any better Via) to replace the Creative Labs, may have heard a 0.03db difference, according to Rightmark Audio Analyzer.
Well, I did.
That was surprising.
Perhaps if it hadn't been an extremely annoying 0.03db, then I wouldn't have detected it at all. 😉 As before, it can matter greatly on what the error is rather than how much it is.
At the time, I just felt deaf, and like no amount of money could get music out of a computer.
As it turns out, two hundred dollars worth of wrong cannot compete with twelve dollars worth of right.
In my opinion, this was related to harmonics because a purely frequency response variation of such a small figure as noticeable, isn't really plausible. I'm not skilled enough to explain harmonics in relation to all of these factors.
I'd like to propose that small, yet obnoxious errors are easily audible; however, large, pleasant errors may pass unnoticed.
Example:
Audio with large errors in favor with the ear can be so exciting as to simultaneously get thousands of Canadians half naked. <that's a link. Click the HD button for decent audio if your internet is reasonable speed. In this example, the bass is "the lower the louder" the voice is "non shout" (she really does sound that way) and the upper treble has an airy lift. If your system isn't capable of system symmetry with this music type, you can plug in some decent headphones to gain reasonable access to the presentation.
And now have a look at this old chart for comparison:
Attachments
Last edited:
Sorry, this makes no sense to me at all. How did you detect the 0.03dB difference? Under what conditions?
I use an MAudio A192 as my measurement card.
I also use the 192 card for both measurement and as a music source. Works great. 0.03dB errors of level or EQ audible? I'd like to see that data.
- Status
- Not open for further replies.